Alberto Romero
2 min readJul 1, 2021


Thank you so much Aniket! It's very kind of you and I appreciate your words :)

"To test a system which is 'sufficiently intelligent'-- pertaining to its capability to 'extrapolate the knowledge that it has learnt' even the system which is to be used for testing these limits, must be 'sufficiently smart' itself." I agree 100% with this. However, I'd qualify the idea by saying that both "sufficient" dot need to refer to exactly the same amount of intelligence.

One sufficient refers to the amount of intelligence something needs to be called intelligent, whereas the other refers to the amount of intelligence something needs to measure another intelligence.

I believe the amount of intelligence required to test another intelligence depends greatly on the intelligence being measured. For instance, a human can't satisfactorily measure the intelligence of another human, although both are equal in intelligence. A dog can't measure the intelligence of an ant despite being more intelligent by a large margin.

Therefore, we can conclude that there's a point above our intelligence where both "sufficients" align. A being that has this amount of intelligence could measure at most, its exact amount of intelligence. However, and this is the key point here, we may always remain below that point. And we could ponder if there's a point above that where an even higher being could measure everything measurable, even those intelligences that are beyond itself.

My point is, there's nothing especial about humans. Our cognoscitive, biological, physical, etc. capabilities are very limited. Nothing assures we could ever measure an intelligence equal or higher than us.

(I may write an article explaining this idea better!)

Again, thank you Aniket for your comment. I really appreciate your kind words :)